ACM SIGMOD Record · 2004 · 14 citations · 17 references
ComputingComputer ScienceInformation SystemsData ManagementDatabasesMachine LearningData IntegrationData ModelingDatabase ManagementIntegration ArchitectureKnowledge IntegrationQuery ProcessingAutomated Query GenerationCross-source Data LinkingData Integration SystemsInformation Integration
We describe three generations of information integration systems developed at George Mason University. All three systems adopt a virtual database design: a global integration schema, a mapping between this schema and the schemas of the participating information sources, and automatic interpretation of global queries. The focus of Multiplex is rapid integration of very large, evolving, and heterogeneous collections of information sources. Fusionplex strengthens these capabilities with powerful tools for resolving data inconsistencies. Finally, Autoplex takes a more proactive approach to integration, by "recruiting" contributions to the global integration schema from available information sources. Using machine learning techniques it confronts a major cost of integration, that of mapping new sources into the global schema.
17
Answering queries using views: A survey
Alon Halevy · The VLDB Journal · 2001 · 1.5K citations
Generic Schema Matching with Cupid
Jayant Madhavan, Philip A. Bernstein, Erhard Rahm · 2001 · 1.3K citations · Full text
Reconciling schemas of disparate data sources
AnHai Doan, Pedro Domingos, Alon Halevy · 2001 · 733 citations